Deep visual unsupervised domain adaptation for classification tasks: a survey

Research output: Contribution to journalArticlepeer-review

Electronic versions

  • Yeganeh Madadi
    Azad University
  • Vahid Seydi
    Azad University
  • Kamal Nasrollahi
    Aalborg University
  • Reshad Hosseini
    University of Tehran
  • Thomas B Moeslund
    Aalborg University
Learning methods are challenged when there is not enough labelled data. It gets worse when the existing learning data have different distributions in different domains. To deal with such situations, deep unsupervised domain adaptation techniques have newly been widely used. This study surveys such domain adaptation methods that have been used for classification tasks in computer vision. The survey includes the very recent papers on this topic that have not been included in the previous surveys and introduces a taxonomy by grouping methods published on unsupervised domain adaptation into five groups of discrepancy-, adversarial-, reconstruction-, representation-, and attention-based methods.
Original languageUnknown
Pages (from-to)3283 – 3299
JournalIET Image Processing
Issue number14
Early online date3 Nov 2020
Publication statusPublished - 1 Dec 2020
Externally publishedYes
View graph of relations